readme changes, versioning

.gitignore

@@ -3,6 +3,8 @@
 ## Ignore Visual Studio temporary files, build results, and
 ## files generated by popular Visual Studio add-ons.
 
+.DS_Store
+
 # User-specific files
 *.suo
 *.user

README.rst

@@ -1,6 +1,10 @@
 PySIMD
 ======
 
+.. image:: images/pysimd.png
+  :width: 800
+  :alt: PySimd logo
+
 ``simd`` is the python module for SIMD computing and programming. It
 prodives an extensive interface to SIMD instruction sets on several different 
 architectures, and fallback scalar implementations when no SIMD instructions 
@@ -129,3 +133,26 @@ The data inside a vector can also be retrieved as a collection type, like a ``tu
 
 The above example shows the pure ``__repr__`` method of ``Vec`` only depicts a hexadecimal, byte level representation of the vector data, but a method like ``as_tuple`` allows the viewing of data with different types. One unique aspect of the ``simd`` module is it treats data and memory similar to that of C, where a chunk of 16 bytes could be two 64 bit integers, four 32 bit integers, and so on.
 
+
+Math
+~~~~
+
+The ``simd`` module supports simd operations that involve artihmetic and math on integers and floating point numbers. Operations like ``add`` or ``sub`` work off another vector and a ``width``. The ``width`` indicates the width of the data lane the simd instruction applies to, such as ``8`` for 64 bit operations. Here are a few examples:
+
+.. code:: py
+
+    >>> v = simd.Vec(size=16, repeat_value=5, repeat_size=4)
+    >>> v2 = simd.Vec(size=16, repeat_value=10, repeat_size=4)
+    >>> v.add(v2)
+    Traceback (most recent call last):
+      File "<stdin>", line 1, in <module>
+    TypeError: function missing required argument 'width' (pos 2)
+    >>> v.add(v2, width=4)
+    >>> v.as_tuple(type=int, width=4)
+    (15, 15, 15, 15)
+    >>> v.sub(v2, width=4)
+    >>> v.sub(v2, width=4)
+    >>> v.as_tuple(type=int, width=4)
+    (-5, -5, -5, -5)
+
+

include/simd_vec_filter.h

@@ -4,6 +4,7 @@
 #include "simd_vec_type.h"
 #include "vec_macros.h"
 
+
 static int pysimd_vec_filter_32(struct pysimd_vec_t* vec, int* gt,
 	                                                      int* lt,
 	                                                      int* eq) {
@@ -118,4 +119,118 @@ static int pysimd_vec_filter_32(struct pysimd_vec_t* vec, int* gt,
 	return 1;
 }
 
+static int pysimd_vec_filter_64(struct pysimd_vec_t* vec, long long* gt,
+	                                                      long long* lt,
+	                                                      long long* eq) {
+#if defined(PYSIMD_X86_SSE2)
+	// pass
+	unsigned char* ptr = vec->data;
+	const unsigned char* ptr_end = ptr + vec->size;
+	while (ptr < ptr_end) {
+		__m128i loaded = _mm_load_si128((__m128i const*)ptr);
+		__m128i mask = _mm_set1_epi8(0xff);
+		if (gt != NULL) {
+			__m128i gtnum = _mm_set1_epi64(*gt);
+			__m128i gtres = _mm_cmpgt_epi64(loaded, gtnum);
+			mask = _mm_and_si128(mask, gtres);
+		}
+		if (lt != NULL) {
+			__m128i ltnum = _mm_set1_epi64(*lt);
+			__m128i ltres = _mm_cmplt_epi64(loaded, ltnum);
+			mask = _mm_and_si128(mask, ltres);
+		}
+		if (eq != NULL) {
+			__m128i eqnum = _mm_set1_epi64(*eq);
+			__m128i eqres = _mm_cmpeq_epi64(loaded, eqnum);
+			mask = _mm_and_si128(mask, eqres);
+		}
+		__m128i final_result = _mm_and_si128(mask, loaded);
+		int mask_result = _mm_movemask_epi8 (final_result);
+		size_t to_advance = 0;
+		switch (mask_result) {
+			case 0xFFFF:
+			case 0x0FFF:
+			case 0x00FF:
+			case 0x000F:
+			case 0x0:
+			      // no filtering needed
+			      to_advance = 16;
+			      break;
+			case 0xFF0F:
+			      final_result = _mm_shuffle_epi32(final_result, 0x78);
+			      to_advance = 12;
+			      break;
+			case 0xF00F:
+			      // shuffle, reverse order of 0b10101100
+			      final_result = _mm_shuffle_epi32(final_result, 0xac);
+			      to_advance = 8;
+			      break;
+			case 0x0FF0:
+			      // shuffle, reverse order of 0b11001001
+			      final_result = _mm_shuffle_epi32(final_result, 0xc9);
+			      to_advance = 8;
+			      break;
+			case 0xF0F0:
+			      // shuffle, reverse order of 0b10001101
+			      final_result = _mm_shuffle_epi32(final_result, 0x8d);
+			      to_advance = 8;
+			      break;
+			case 0x0F0F:
+			      // shuffle, reverse order of 0b11011000
+			      final_result = _mm_shuffle_epi32(final_result, 0xd8);
+			      to_advance = 8;
+			      break;
+			case 0xF0FF:
+			      final_result = _mm_shuffle_epi32(final_result, 0xb4);
+			      to_advance = 12;
+			      break;
+			case 0xFFF0:
+			      // shuffle, reversed order of 00011011
+			      final_result = _mm_shuffle_epi32(final_result, 0x1b);
+			      to_advance = 12;
+			      break;
+			case 0xFF00:
+			      // shuffle, reversed order of 00001110
+			      final_result = _mm_shuffle_epi32(final_result, 0xe);
+			      to_advance = 8;
+			      break;
+			case 0xF000:
+			      // shuffle, reversed order of 00000011
+			      final_result = _mm_shuffle_epi32(final_result, 0x3);
+			      to_advance = 4;
+			      break;
+			case 0x0F00:
+			      // shuffle, reversed order of 0b00000010
+			      final_result = _mm_shuffle_epi32(final_result, 0x2);
+			      to_advance = 4;
+			      break;
+			case 0x00F0:
+			      // shuffle, reversed order of 0b00000001
+			      final_result = _mm_shuffle_epi32(final_result, 0x1);
+			      to_advance = 4;
+			      break;
+			default:
+			      fprintf(stderr, "Got impossible mask value: 0x%x, aborting ...\n", mask_result);
+			      abort();
+		}
+		_mm_store_si128 ((__m128i*)ptr, final_result);
+		ptr += to_advance;
+	}
+#else
+	const unsigned char* reader = vec->data;
+	const unsigned char* read_end = reader + vec->size;
+	void* new_buf = calloc(1, vec->size);
+	unsigned char* writer = new_buf;
+	while (reader < read_end) {
+		if (*reader) {
+			*writer++ = *reader;
+		}
+		++reader;
+	}
+	free(vec->data);
+	vec->data = new_buf;
+#endif
+	return 1;
+}
+
 #endif // PYSIMD_VEC_FILTER_H

setup.py

@@ -10,8 +10,8 @@
 # on vector object without needing to check the length/size of it
 pysimd_minimum_align = 8
 
-pysimd_patch_version = 4
-pysimd_minor_version = 0
+pysimd_patch_version = 0
+pysimd_minor_version = 1
 pysimd_major_version = 0
 
 pysimd_version = [pysimd_major_version, 

src/pymain.c

@@ -1,7 +1,7 @@
 #include "core_simd_info.h"
 #include "simd_vec.h"
 #include "simd_vec_arith.h"
-#include "simd_vec_filter.h"
+//#include "simd_vec_filter.h"
 #define PY_SSIZE_T_CLEAN
 #include <Python.h>
 #include "structmember.h"